setwd("/Users/jaeheehwang/Documents/gelman/")
# Import Data
library(plotly)
## Warning: package 'plotly' was built under R version 3.2.3
## Loading required package: ggplot2
## Warning: package 'ggplot2' was built under R version 3.2.3
##
## Attaching package: 'plotly'
##
## The following object is masked from 'package:ggplot2':
##
## last_plot
##
## The following object is masked from 'package:graphics':
##
## layout
library(ggplot2)
library(package = "lattice")
library(ggmap)
##
## Attaching package: 'ggmap'
##
## The following object is masked from 'package:plotly':
##
## wind
library(source.gist)
## Loading required package: RCurl
## Loading required package: bitops
## Loading required package: rjson
library(rworldmap)
## Warning: package 'rworldmap' was built under R version 3.2.3
## Loading required package: sp
## ### Welcome to rworldmap ###
## For a short introduction type : vignette('rworldmap')
library("Hmisc")
## Warning: package 'Hmisc' was built under R version 3.2.3
## Loading required package: survival
## Loading required package: Formula
##
## Attaching package: 'Hmisc'
##
## The following object is masked from 'package:plotly':
##
## subplot
##
## The following objects are masked from 'package:base':
##
## format.pval, round.POSIXt, trunc.POSIXt, units
library(base)
dat <- read.csv ("cleanwithGDP.csv",header=TRUE, stringsAsFactors=FALSE, fileEncoding="latin1")
datc <- dat
variable.names(datc)
## [1] "registerNum" "country1" "country2"
## [4] "dateBegan" "dateEnded" "durationDays"
## [7] "peopleDead" "peopleDisplaced" "damageUsd"
## [10] "mainCause1" "mainCause2" "mainCause3"
## [13] "severity" "affectedSqKm" "magnitude"
## [16] "centroidX" "centroidY" "GDPmil"
## [19] "GDPrank" "Income.Group.OECD" "Region"
## [22] "Land.Area"
########################################################################
#### AREA AFFECTED
# Let's see how the distribution of affected sqkm range.
areaInterval <- cut2(datc$affectedSqKm, c(10000,50000,100000,10000000))
table(areaInterval)
## areaInterval
## [1.2e+01,1.0e+04) [1.0e+04,5.0e+04) [5.0e+04,1.0e+05) [1.0e+05,1.0e+07]
## 1196 1395 622 1099
# Let's divide affected area into small, medium, large, or very large intervals.
datc$affected <- ifelse(datc$affectedSqKm < 10000, "Small", ifelse(datc$affectedSqKm < 50000, "Medium", ifelse(datc$affectedSqKm < 100000, "Large", "Very Large")))
datc$affected = factor(datc$affected,levels=c("Small", "Medium", "Large", "Very Large"), ordered = TRUE)
table(datc$affected)
##
## Small Medium Large Very Large
## 1196 1395 622 1099
########################################################################
#### People Dead
table(cut2(datc$peopleDead, g = 10))
##
## 0 1 [ 2, 4) [ 4, 6) [ 6, 11) [11, 19)
## 1231 235 420 298 435 405
## [19, 36) [36, 88) [88,160000]
## 444 413 431
deadInterval <- cut2(datc$peopleDead, c(10,50,100,10000000))
table(deadInterval)
## deadInterval
## [0e+00,1e+01) [1e+01,5e+01) [5e+01,1e+02) [1e+02,1e+07]
## 2549 1093 282 388
# Let's divide people dead into intervals
datc$dead <- ifelse(datc$peopleDead < 10, "Under 10", ifelse(datc$peopleDead < 50, "10-50", ifelse(datc$peopleDead < 100, "50-100", "Over 100")))
datc$dead = factor(datc$dead,levels=c("Under 10", "10-50","50-100","Over 100"), ordered = TRUE)
table(datc$dead)
##
## Under 10 10-50 50-100 Over 100
## 2549 1093 282 388
########################################################################
#### People displaced
table(cut2(datc$peopleDisplaced, g = 10))
##
## 0 [ 2, 21) [ 21, 406) [ 406, 1600)
## 1278 19 465 414
## [ 1600, 4080) [ 4080, 10308) [ 10308, 31000) [ 31000, 150000)
## 461 389 426 429
## [150000,40000000]
## 431
displacedInterval <- cut2(datc$peopleDisplaced, c(30,5000,100000,100000000))
table(displacedInterval)
## displacedInterval
## [0e+00,3e+01) [3e+01,5e+03) [5e+03,1e+05) [1e+05,1e+08]
## 1306 1356 1114 536
# Let's divide people displaced into intervals
datc$displaced <- ifelse(datc$peopleDisplaced < 30, "Under 30", ifelse(datc$peopleDisplaced < 5000, "30-5000", ifelse(datc$peopleDisplaced < 100000, "5000-100000", "Over 100000")))
table(datc$displaced)
##
## 30-5000 5000-100000 Over 100000 Under 30
## 1356 1114 536 1306
datc$displaced = factor(datc$displaced,levels=c("Under 30", "30-5000","5000-100000","Over 100000"), ordered = TRUE)
########################################################################
#### Duration
table(cut2(datc$durationDays, g = 10))
##
## [ 1, 3) 3 4 5 6 [ 7, 9) [ 9, 11) [11, 16)
## 587 573 486 427 262 435 289 444
## [16, 26) [26,419]
## 398 411
durationInterval <- cut2(datc$durationDays, c(5, 10, 20, 500))
table(durationInterval)
## durationInterval
## [ 1, 5) [ 5, 10) [ 10, 20) [ 20,500]
## 1646 1278 776 612
# Let's divide duration into intervals
datc$duration <- ifelse(datc$durationDays < 5, "Under 5", ifelse(datc$durationDays < 10, "5-10", ifelse(datc$durationDays < 20, "10-20", "Over 20")))
table(datc$duration)
##
## 10-20 5-10 Over 20 Under 5
## 776 1278 612 1646
datc$duration = factor(datc$duration,levels=c("Under 5", "5-10","10-20","Over 20"), ordered = TRUE)
#########################################################################
table(datc$Income.Group.OECD)
##
## High income: nonOECD High income: OECD
## 100 203 990
## Low income Lower middle income Upper middle income
## 452 1357 1210
datc$incomeGroup <- datc$Income.Group.OECD
datc$incomeGroup = factor(datc$incomeGroup, levels = c("Low income", "Lower middle income", "Upper middle income","High income: nonOECD", "High income: OECD"), ordered = TRUE)
table(datc$incomeGroup)
##
## Low income Lower middle income Upper middle income
## 452 1357 1210
## High income: nonOECD High income: OECD
## 203 990
datc$GDPlevel <- ifelse(datc$Income.Group.OECD == "High income: OECD", 5,
ifelse(datc$Income.Group.OECD == "High income: nonOECD", 4,
ifelse(datc$Income.Group.OECD == "Upper middle income", 3,
ifelse(datc$Income.Group.OECD == "Lower middle income", 2, 1))))
table(datc$GDPlevel)
##
## 1 2 3 4 5
## 552 1357 1210 203 990
#########################################################################
#########################################################################
#########################################################################
#########################################################################
#########################################################################
#plots
#GDP rank by region
#North America has high GDP. Sub-saharan africa seems to have low GDP.
bwplot(~datc$GDPrank|factor(datc$Region),data=datc,col="blue",
main="GDP rank by region")
#income group by region
#North America has many high income group countries, whereas south asia and sub-saharan africa has many low-income countries.
bwplot(~datc$incomeGroup|factor(datc$Region),data=datc,col="blue",
main="income group by region")
# People displacement and country income group.
dotplot(datc$peopleDisplaced ~ datc$magnitude | datc$incomeGroup, data = datc)
dotplot(datc$displaced ~ datc$magnitude | datc$incomeGroup, data = datc)
# People dead and country income group.
dotplot(datc$peopleDead ~ datc$magnitude | datc$incomeGroup, data = datc)
dotplot(datc$dead ~ datc$magnitude | datc$incomeGroup, data = datc)
# Magnitude and affected square km are positively correlated in a S shaped curve.
# dotplot(datc$affectedSqKm ~ datc$magnitude | datc$Region, data = datc)
dotplot(datc$affected ~ datc$magnitude | datc$Region, data = datc)
# In North America Europe, Middle East and North Africa,
# less people were likely to be dead regardless of magnitude and squaremeter affected
# In South Africa, more people were likely to be dead when affected squaremeter was larger.
dotplot(datc$affectedSqKm ~ datc$dead | datc$Region, data = datc)
dotplot(datc$affectedSqKm ~ datc$displaced | datc$Region, data = datc)
dotplot(datc$magnitude ~ datc$dead | datc$Region, data = datc)
dotplot(datc$magnitude ~ datc$displaced | datc$Region, data = datc)
dotplot(datc$magnitude ~ datc$dead | datc$incomeGroup, data = datc)
dotplot(datc$magnitude ~ datc$displaced | datc$incomeGroup, data = datc)
# For high income countries, less people died regardless of affected squaremeter or magnitude.
# For less than high income countries, more people were prone to die at higher squarementer and magnitude.
dotplot(datc$affectedSqKm ~ datc$dead | datc$incomeGroup, data = datc)
dotplot(datc$affectedSqKm ~ datc$displaced | datc$incomeGroup, data = datc)
bwplot(~datc$displaced|factor(datc$incomeGroup),data=datc,col="blue",
main="Displacement by Country")
bwplot(~datc$dead|factor(datc$incomeGroup),data=datc,col="blue",
main="Number Dead by Country")
dotplot(datc$dead ~ datc$GDPrank | datc$Region, data = datc)
dotplot(datc$affected ~ datc$GDPrank | datc$Region, data = datc)
#magnitude and affected area by region
#monsoons seem to bring high magnitude floods.
bwplot(~datc$magnitude|factor(datc$mainCause1),data=datc,col="blue",
main="magnitude by main cause")
#monsoons seem to affect large regions
#landslides seem to affect small regions
#tsunami seems to affect wide array of regions.
bwplot(~datc$affected|factor(datc$mainCause1),data=datc,col="blue",
main="affected area by main cause")
# tsunamis seem to kill many people.
#snow/icemelt seem to usually not kill many.
bwplot(~datc$dead|factor(datc$mainCause1),data=datc,col="blue",
main="death by main cause")
#hurricanes and monsoons seem to displace a lot of people.
#landslide and avalanches seem to now displace many.
bwplot(~datc$displaced|factor(datc$mainCause1),data=datc,col="blue",
main="displacement by main cause")
dotplot(datc$damageUsd ~ datc$incomeGroup | datc$Region, data = datc)
#main causes by region.
#stacked histogram. diff lengths.
#East Asia seems to have predominantly many floods.
hist <- ggplot(datc, aes(x = datc$Region, fill = datc$mainCause1))
hist + geom_bar(color = "black", aes(fill = datc$mainCause1, position = "fill"))+
labs(title = "Main Causes by Region", x = "Region", y = "Percent") +
scale_fill_discrete(name="Main Causes") + guides(fill = guide_legend(reverse=TRUE))
#equal height
#Main cause seems to be Heavy Rain, except for South Asia where Monsoon also seems to be a leading cause.
#East Asia & Pacific also had some Trophical Storms.
hist2 <- ggplot(datc, aes(x = datc$Region))
hist + geom_bar(color = "black", aes(fill = datc$mainCause1), position = "fill")+
labs(title = "Main Causes by Region", x = "Region", y = "Percent") +
scale_fill_discrete(name="Main Causes") + guides(fill = guide_legend(reverse=TRUE))
###########################################################################################
# other ideas: radar chart?
# http://www.cmap.polytechnique.fr/~lepennec/R/Radar/RadarAndParallelPlots.html
# can we throw in a bubble graph?
###########################################################################################
#mapping
library(plotly)
library(ggplot2)
library(package = "lattice")
library(ggmap)
library(source.gist)
library(rworldmap)
dat <- read.csv ("cleanwithGDP.csv",header=TRUE, stringsAsFactors=FALSE, fileEncoding="latin1")
datc <- dat
variable.names(datc)
## [1] "registerNum" "country1" "country2"
## [4] "dateBegan" "dateEnded" "durationDays"
## [7] "peopleDead" "peopleDisplaced" "damageUsd"
## [10] "mainCause1" "mainCause2" "mainCause3"
## [13] "severity" "affectedSqKm" "magnitude"
## [16] "centroidX" "centroidY" "GDPmil"
## [19] "GDPrank" "Income.Group.OECD" "Region"
## [22] "Land.Area"
library(rworldmap)
library(maps)
## Warning: package 'maps' was built under R version 3.2.3
##
## # ATTENTION: maps v3.0 has an updated 'world' map. #
## # Many country borders and names have changed since 1990. #
## # Type '?world' or 'news(package="maps")'. See README_v3. #
library(ggplot2)
world_map <- map_data("world")
p <- ggplot() + coord_fixed() + xlab("") + ylab("")
base_world_messy <- p + geom_polygon(data=world_map, aes(x=long, y=lat, group=group),
colour="light green", fill="light green")
base_world_messy
cleanup <-
theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(),
panel.background = element_rect(fill = 'white', colour = 'white'),
axis.line = element_line(colour = "white"), legend.position="none",
axis.ticks=element_blank(), axis.text.x=element_blank(),
axis.text.y=element_blank())
datc1 <- datc[1:500,]
base_world <- base_world_messy + cleanup
map_data <-
base_world +
geom_point(data=datc1,
aes(x=centroidX, y=centroidY), colour="Deep Pink",
fill="Pink",pch=21, size=5, alpha=I(0.7))
variable.names(datc1)
## [1] "registerNum" "country1" "country2"
## [4] "dateBegan" "dateEnded" "durationDays"
## [7] "peopleDead" "peopleDisplaced" "damageUsd"
## [10] "mainCause1" "mainCause2" "mainCause3"
## [13] "severity" "affectedSqKm" "magnitude"
## [16] "centroidX" "centroidY" "GDPmil"
## [19] "GDPrank" "Income.Group.OECD" "Region"
## [22] "Land.Area"
datc1$Income.Group <- datc1$Income.Group.OECD
#map_data_coloured <- base_world +
# geom_point(data=datc1,
# aes(x=centroidX, y=centroidY, colour=datc$Income.Group), size=3, alpha=I(0.7))
#map_data_coloured
#flood causes and affected sqkm
map_data_sized <-
base_world +
geom_point(data=datc1,
aes(x=centroidX, y=centroidY, colour=mainCause1, size=affectedSqKm), alpha=I(0.7))
map_data_sized
#magnitude and affectedsqkm
MagAndSqkm <-
base_world +
geom_point(data=datc1,
aes(x=centroidX, y=centroidY, colour=magnitude, size=affectedSqKm), alpha=I(0.7))
MagAndSqkm
map_data_sized <-
base_world +
geom_point(data=datc1,
aes(x=centroidX, y=centroidY, colour=mainCause1, size=damageUsd), alpha=I(0.7))
map_data_sized
###########################################################################################
###########################################################################################
###########################################################################################
# other ideas: radar chart?
# http://www.cmap.polytechnique.fr/~lepennec/R/Radar/RadarAndParallelPlots.html
# can we throw in a bubble graph?
###########################################################################################